The goal of our work was to develop a mass-spectrometry-based proteomics workflow to quantify TSST-1, SEA, SEC, and SED in menstrual fluids to allow us to explore the pathophysiology of mTSS.

Recent interest in non-invasive diagnostic approaches to address women’s health issues, including infertility and uterine pathologies, has led to a search for methods to characterize menstrual fluid [22]. MS-based proteomics analysis can provide relevant information on the protein elements contained in menstrual fluids, which could be useful for histopathological research and for diagnostic purposes [10]. However, menstrual fluid is difficult to analyze due to its complexity and variations in its protein content. Consequently, very few articles have been published describing MS-based pathophysiology studies or assays to detect protein biomarkers [22]. Here, we harnessed the power of targeted MS-based proteomics to investigate the presence of staphylococcal superantigens produced in menstrual fluids from patients with mTSS, using menstrual fluid from healthy women as the control. Based on our previous experience in detecting SEs and TSST-1 in complex biological samples including serum [23], urine [19], and food [24], we developed a targeted proteomics workflow combining efficient biochemical preparation and quantitative LC-SRM to analyze menstrual fluids extracted from tampons used during menstruation. As expected, pre-analytical optimizations and analytical developments were challenging due to the variable composition of menstrual fluids, similar to food samples (Supplementary Figure S2) [24]. In addition, the presence of red blood cells, serum, uterine tissue, and vaginal secretions in the menstrual fluid extended the dynamic range of protein abundance beyond the 12 orders of magnitude described for serum or plasma matrices [11]. However, the combination of an efficient digestion protocol with an optimized LC-SRM analytical method ensured the sensitive detection of the toxins in clinical samples extracted from tampons. Our results confirmed the presence of TSST-1—at concentrations ranging between 10 ng/mL and 460 ng/mL—in menstrual fluid collected from 5 of the 6 women with mTSS. These results can be compared to two previous reports of vaginal detection of TSST-1 in mTSS. In 1987, Rosten and coworkers [16] developed an enzyme-linked immunosorbent assay (ELISA) to detect TSST-1 in vaginal washings. Their assay detected TSST-1 in clinical samples from 2 out of 9 patients with mTSS, only during the acute phase, with concentrations ranging from 2.2 ng/mL to 15.8 ng/mL. These results appear compatible with our results as vaginal washings are likely to be more diluted than menstrual fluids. In 2010, Schlievert and coworkers [25] used semi-quantitative western-blotting and compared signal intensities to TSST-1 standard curves to investigate the presence of TSST-1 in tampons collected from two women with mTSS. For these two patients, they reported TSST-1 levels of 69 and 80 µg per tampon (estimated tampon volume, 3 to 10 mL). These estimations are 20-fold higher than our results. It should be noted that these authors provided no information on the method used to quantify the purified TSST-1 used to create the standard curve. Our quantification results were obtained using AAA-calibrated isotopically-labeled standards, which have been demonstrated to be highly accurate. Consequently, differences in accuracy of the reference levels could explain the apparent discrepancies between the concentrations measured. In accordance with the results presented in these previous studies, we also detected TSST-1 in a sample from one healthy woman host to vaginal tst+ S. aureus. Thus, TSST-1 can be produced by S. aureus in the vagina during menstruation in the presence of a tampon without necessarily leading to mTSS. The pathophysiology of mTSS is complex [3]. The different steps for mTSS occurrence are the production of TSST-1 by S. aureus in the vaginal fluid, the passage of the toxin through the vaginal mucosa, and the activation of the immune system. In addition, natural antibodies neutralizing the toxin must be absent. MS-based detection of TSST-1 in menstrual fluids explores the first step of mTSS pathophysiology only. This explains why the detection of toxins in the vagina alone, as observed in one healthy woman, does not predict the occurrence of the disease. However, in our population, deliberately biased to include women with vaginal colonization by S. aureus, the detection of TSST-1 in the vaginal fluid was statistically well correlated with the occurrence of mTSS.The level of TSST-1 production varied between women, underlining the importance of the appropriate use of intravaginal devices. No correlation was detected between the amount of TSST-1 detected in the vaginal fluid and the clinical or microbiological characteristics of the patients with mTSS. Interestingly, in one patient with mTSS, TSST-1 was undetectable by our method. There are numerous proteases within menstrual fluids [10] and protease inhibitors were not added to the menstrual fluids before storage. Possibly, toxins may have been degraded by proteases, especially TSST-1, which is less resistant to proteolysis than enterotoxins [10]. Based on this hypothesis, TSST-1 concentrations may be underestimated.An issue that is not yet fully resolved is the almost exclusive association between TSST-1 secretion and the menstrual form of toxic shock [7]. Thanks to the multiplexing capabilities of targeted proteomics, we were able to simultaneously investigate the production of other staphylococcal superantigenic toxins in the vagina—including SEA, SEC, and SED—that have been associated with non-mTSS. In our cohort, SEA and SED were never detected in menstrual fluids, whereas SEC was detected in two clinical specimens, one collected from a patient with mTSS and one from a healthy volunteer whose S. aureus isolate was positive for tst, sec, and sed. Although co-secretion of TSST-1 could not be confirmed (due to signal contamination inducing changes in the relative intensity of SRM transitions), these results indicate that both TSST-1 and SEC can be produced by S. aureus in menstrual fluids, challenging the hypothesis that the epidemiological link between mTSS and the exclusive production of TSST-1 among staphylococcal superantigens is related to local conditions in the vagina, specifically inducing TSST-1 expression [1,26].

In the future, we plan to continue using the method presented here to explore toxin production in vivo in users of intravaginal devices to improve our understanding of the pathophysiology of mTSS and its prevention.